Parkinson's Disease (PD) motor rehabilitation can improve clinical balance and disease severity measures. However, clinical scales cannot identify changes in specific neural pathways, and therefore may not be able to discriminate between remediative and compensatory changes to underlying neural mechanisms during rehabilitation. Our long-term goal is to discriminate between remediation and compensation during effective PD rehabilitation, which would significantly improve our understanding of response to treatment in individuals with PD as well as potentially in other neurological conditions. In contrast to clinical measures, we will study neuromuscular activation measured with electromyography (EMG) as a direct window into the neuromotor activity of individuals with PD before and after 24 weeks of adapted tango rehabilitative dance classes, which is an effective rehabilitative intervention with low attrition. We will quantify neuromotor deficits and recovery before and after rehabilitation using our sensorimotor response model (SRM), which was developed to quantify brainstem-mediated EMG responses evoked during perturbations to standing balance in young individuals as well as in animal models. We will use a novel experimental paradigm to determine whether motor improvements associated with rehabilitation are due to remediative changes in basal ganglia-brainstem pathways or due to compensatory changes in cerebellar-brainstem pathways. Previous studies of perturbation responses during standing balance demonstrate that basal ganglia pathways are used to adapt neural control parameters to new contexts immediately, whereas cerebellar pathways are used to adapt to new contexts slowly, over repeated trials. Therefore, we will examine the timecourse of adaptation of SRM parameters over repeated perturbation trials in novel stance widths. We predict that immediate changes in SRM parameters in novel stance widths will reflect basal ganglia-brainstem remediation, whereas slowly adapting changes will reflect cerebellar compensation. After rehabilitation, we predict that increased magnitudes of immediate changes will reflect basal ganglia-brainstem remediation, whereas increased speed of slowly adapting changes will reflect cerebellar-brainstem compensation. Further, we predict that SRM parameters can serve as biomarkers to identify differences in pathways underlying motor deficits across participants with PD throughout rehabilitation.